Show all RowNumber records for duplicates? - sql-server

Question

I have a table of string resources :
;WITH cte AS
(
SELECT 1 AS id , 'john' AS name, 10 AS age
UNION
SELECT 2 AS id , 'john' AS name, 10 AS age
UNION
SELECT 3 AS id , 'john' AS name, 12 AS age
UNION
SELECT 4 AS id , 'paul' AS name, 6 AS age
UNION
SELECT 5 AS id , 'paul ' AS name, 6 AS age
UNION
SELECT 6 AS id , 'paul different' AS name, 7 AS age
UNION
SELECT 7 AS id , 'ringo' AS name, 2 AS age
)
So the name "john" has age of 10.
Later on , someone else (not me) also added "john" with age 10.
So I want to clean all duplicates.
But that's not the problem. Before I delete I want to see all duplicates.
So I did this :
SELECT *
FROM (
SELECT ID,
name,
age,
ROW_NUMBER() OVER(PARTITION BY name, age ORDER BY id) AS rn
FROM cte
) a WHERE a.rn>1
ORDER BY
name,
age,
a.rn
Result :
Which basically shows me duplicates. But I want to see also where rn=1 only if there's more version for the current value.
Question
In other words : How can I enhance my query so :
Show all versions for a record ( all row numbers , rn) only if there are versions for this record
Desired result :
ID name age rn
1 john 10 1
2 john 10 2
4 paul 6 1
5 paul 6 2
Sql online - demo
NB I know i can do it with rescanning the table for the same name and age . bUt I thought if there's more elegant way of doing it.

Answer 1

Use exists operator to find the name which are duplicated. Try this.
;WITH cte AS
(
SELECT 1 AS id , 'john' AS name, 10 AS age
UNION
SELECT 2 AS id , 'john' AS name, 10 AS age
UNION
SELECT 3 AS id , 'john' AS name, 12 AS age
UNION
SELECT 4 AS id , 'paul' AS name, 6 AS age
UNION
SELECT 5 AS id , 'paul ' AS name, 6 AS age
UNION
SELECT 6 AS id , 'paul different' AS name, 7 AS age
UNION
SELECT 7 AS id , 'ringo' AS name, 2 AS age
)
, cte1
AS (SELECT ID,
name,
age,
Row_number() OVER(PARTITION BY name, age ORDER BY id) AS rn
FROM cte)
SELECT *
FROM cte1 a
WHERE EXISTS (SELECT 1
FROM cte1 b
WHERE a.name = b.name and a.age=b.age
AND b.rn > 1)
ORDER BY name, age, a.rn
or use Inner Join
SELECT a.id,a.name,a.age
FROM cte1 a
JOIN cte1 b
ON a.name = b.name
AND a.age = b.age
AND b.rn > 1
ORDER BY a.name, a.age, a.rn
Or To do it in single table scan use Dense_Rank plus window function
;WITH cte AS
(
SELECT 1 AS id , 'john' AS name, 10 AS age
UNION
SELECT 2 AS id , 'john' AS name, 10 AS age
UNION
SELECT 3 AS id , 'john' AS name, 12 AS age
UNION
SELECT 4 AS id , 'paul' AS name, 6 AS age
UNION
SELECT 5 AS id , 'paul ' AS name, 6 AS age
UNION
SELECT 6 AS id , 'paul different' AS name, 7 AS age
UNION
SELECT 7 AS id , 'ringo' AS name, 2 AS age
)
, cte1
AS (SELECT ID,
name,
age,
count(age) over (partition by name,age) cnt,
dense_rank() OVER(PARTITION BY name ORDER BY age) AS rn
FROM cte)
SELECT *
FROM cte1
WHERE rn = 1
AND cnt > 1